Online health community experiences of sexual minority women with cancer

Abstract

Objective

The study sought to explore online health communities (OHCs) for sexual minority women (SMW) with cancer by conducting computational text analysis on posts.

Materials and Methods

Eight moderated OHCs were hosted by the National LGBT Cancer Network from 2013 to 2015. Forty-six SMW wrote a total of 885 posts across the OHCs, which were analyzed using Linguistic Inquiry and Word Count and latent Dirichlet allocation. Pearson correlation was calculated between Linguistic Inquiry and Word Count word categories and participant engagement in the OHCs. Latent Dirichlet allocation was used to derive main topics.

Results

Participants (average age 46 years; 89% white/non-Hispanic) who used more sadness, female-reference, drives, and religion-related words were more likely to post in the OHCs. Ten topics emerged: coping, holidays and vacation, cancer diagnosis and treatment, structure of day-to-day life, self-care, loved ones, physical recovery, support systems, body image, and symptom management. Coping was the most common topic; symptom management was the least common topic.

Discussion

Highly engaged SMW in the OHCs connected to others via their shared female gender identity. Topics discussed in these OHCs were similar to OHCs for heterosexual women, and sexual identity was not a dominant topic. The presence of OHC moderators may have driven participation. Formal comparison between sexual minority and heterosexual women’s OHCs are needed.

Conclusions

Our findings contribute to a better understanding of the experiences of SMW cancer survivors and can inform the development of tailored OHC-based interventions for SMW who are survivors of cancer.

INTRODUCTION

Currently, 15.5 million people are living with cancer in the United States and the number is expected to grow.¹ Diagnosis of and treatment for cancer can result in high levels of psychological distress and an elevated need for emotional and social support among survivors (defined as any person with a history of cancer from time of diagnosis to end of life).^2–5 Moreover, concurrent changes in physical function, body image, and role performance can cause distress throughout cancer treatment and posttreatment.⁶ Overall, many cancer survivors experience clinically significant psychological distress following their diagnosis; while most recover over time, a proportion of survivors (approximately 20%-25%) experience chronic distress.^5–7 The prevalence of and recovery from psychological stress may differ for different subgroups of survivors. Because sexual minority women (SMW) (eg, lesbians, bisexuals, and queer-identified persons) experience a higher rate of mental health diagnoses compared with heterosexual women,⁸ they may be at commensurately higher risk for psychological distress after a diagnosis of cancer.

SMW may experience unique barriers when seeking cancer care that could contribute to their distress and need for support.^9,10 Psychosocial support services that meet the needs of heterosexual women may not address the concerns of SMW with cancer, who consistently request tailored support.^11,12 For example, in a qualitative study, 1 lesbian cancer survivor opted not to participate in a general cancer support group, stating that “[the women in the group] got family, husbands, and I just didn’t feel like I would fit in. I had enough stress on me, I didn’t need to be trying to explain to people why I live the way I live. So that’s why I didn’t do that. Many times, I wish I had somebody to talk to that really understood.”¹³ Unfortunately, very few oncology practices, cancer centers, or community organizations offer services to SMW with cancer that are tailored to their unique needs.

With the rise of the Internet, online health communities (OHCs) have become an increasingly popular means of giving and receiving support for cancer survivors.^14–16 Through engaging social media and sharing health-related information, survivors can become more active participants in their healthcare and play important roles in groups through contributing their experiences and ideas.^17,18 Moreover, OHCs can be tailored to the needs of geographically diffuse minority and underserved populations, including rural residents,^19,20 racial/ethnic minorities,^21,22 and, theoretically, sexual minorities. These tailored OHCs provide improved access to support services and allow minority community members to give and receive support from other cancer survivors who share the same identities.

OHCs have enabled cancer survivors to convey their perspectives or insights about their disease and treatment, even when these perspectives differ from those of their healthcare providers.^18,23,24 Such patient-authored information can be a valuable source of personalized data for both patients, healthcare providers, and researchers, provided it can be systematically analyzed and interpreted. In recent years, an alternative to qualitative analysis, computational text analysis, has gained attention. One example is Linguistic Inquiry and Word Count (LIWC). The LIWC dictionary is a tool that quantifies and categorizes natural language data into categories that reflect linguistic dimensions (eg, pronouns, articles), psychological constructs (eg, affective, cognitive), and personal concerns (eg, work, religion).²⁵ For example, textual items such as email, family, and meeting belong to the social category and love, happy, and nervous belong to the emotion category. Another analytic tool is latent Dirichlet allocation (LDA), a simple topic-modeling algorithm that identifies thematic information from large bodies of text.²⁶ Using LDA, groups of words are parsed from text by an algorithm, and then researchers label these groups of words by the thematic relationship between words. LIWC and LDA have been applied to OHCs used by cancer survivors.^27–31 These methods allow researchers to summarize large bodies of text that would be highly impractical—if not impossible—to annotate by hand.

Leveraging these computational text analyses to understand topics discussed in sexual minority–specific OHCs can shed light on the unique concerns of this understudied population. To date, no studies have examined OHCs specific to SMW with cancer. Therefore, the objective of the study described herein was to explore OHCs for SMW with cancer by conducting computational text analysis (LIWC, LDA) on posts. The results of this study can be used to provide guidance to others interested in both creating and hosting such OHCs for SMW and can be used to design interventions that address community-based concerns.

Materials and METHODS

Participants

The National LGBT Cancer Network (hereafter, the Network) received funding from the New York State Department of Health to host OHCs for sexual minority cancer survivors. Beginning in September 2013, the Network conducted outreach for these OHCs by soliciting potential participants through social media advertisements and mailings to cancer and LGBT organizations. LGBT-identified individuals with any type of cancer were invited to participate; the groups were advertised as being specifically for lesbian, bisexual, and queer-identified women. Interested individuals contacted the Network, and completed a registration form that included demographic information, an item about the previous use of cancer support groups, and characteristics of their cancer (ie, type of cancer, time since diagnosis, and stage).

Individuals were invited to participate in an OHC after completing or at least partially completing the registration form, which meant we have missing demographic information on some participants. The Network conducted 8 OHCs for SMW, comprising a total of 47 participants with any type of cancer. The first OHC was offered for 12 weeks, and the next 7 OHCs were offered for 8 weeks, in response to feedback from the first group of participants. A 2-week break period was scheduled after each OHC was completed, during which the Network conducted outreach for the next OHC. Participants were permitted to participate in as many OHCs as they chose. The Executive Director of the Network (a clinical social worker) and the Network Program Manager co-moderated all 8 OHCs. Because the Program Manager did not have prior group leadership experience, the Executive Director checked the OHC every day and guided the Program Manager to ask open-ended questions/prompts (eg, “Who do you consider part of your support team?”) if several days passed without postings. The software used to facilitate the OHCs allowed for multiple threads within each group. Each thread was based on a single topic, such as Plans for the Holidays. Group members could read and post to all open threads, in addition to starting new threads. Moderators read all posts and occasionally provided comments.

Data

The data for this analysis were the de-identified posts of participants from the 8 OHCs for SMW with cancer. The University of Pittsburgh Institutional Review Board determined this study to be exempt from Institutional Review Board oversight since the posts were analyzed without patient identifiers. While 47 SMW with a diagnosis of cancer were involved in the 8 OHCs, we later excluded 1 participant because they never posted. Therefore, our dataset consisted of 885 posts that were created by 46 participants. The OHC posts were analyzed to understand relevant topics through 2 computational approaches: LIWC and LDA. Because natural language processing of patient-authored information is complex, the use of 2 approaches provided multiple aspects to understand the same set of textual data.

Measures and analysis

Linguistic inquiry and word count (LIWC)

We assessed word usage using the 2015 LIWC dictionary, which includes 4 summary language variables and 88 word categories. We present data on 4 summary variables; the frequency of textual items that fell under each variable was scored and converted to percentiles (based on the area under a normal curve) based on normative text samples tested during the validation of the LIWC program.³²Table 1 describes the summary variables.

Table 1.

Language summary variables of Linguistic Inquiry and Word Count analysis

Summary Variable 32	Measurement32
Analytical thinking	High numbers reflect formal, logical, and hierarchical thinking; lower numbers reflect more informal, personal, here-and-now, and narrative thinking.
Clout	High numbers suggest that the author is speaking from the perspective of high expertise and is confident; lower numbers suggest a more tentative, humble, or even anxious style.
Authenticity	Higher numbers are associated with a more honest, personal, and disclosing text; lower numbers are associated with a more guarded, distanced form of discourse.
Emotional tone	Higher numbers reveal a more positive, upbeat style; lower numbers reveal greater anxiety, sadness, or hostility.

We also calculated 88 additional LIWC categories (for details, see Supplementary Appendix A). We ran an exploratory Pearson correlation analysis between LIWC categories and the OHC engagement rate of SMW in the OHCs, using a .05 significance level. We defined OHC engagement rate as the mean number of postings per participant in each OHC.³³ For the correlation analysis, we limited ourselves to 52 LIWC categories, excluding 36 categories that are nonsemantic (eg, number of words per sentence) or punctuation (eg, commas).

Latent dirichlet allocation (LDA)

The premise of LDA is that each unit of text in a dataset can be represented as a collection of words, and certain groups of words tend to co-occur in text documents. These words represent a mixture of topics, and the topics are characterized by a distribution over a fixed vocabulary. By modeling topics based on probability, one can discover hidden meanings embedded in a unit of text and distinguish between uses of words with multiple meanings.^26,34 To conduct the LDA analysis, we first cleaned our text, excluding nonstandard characters, numbers, common punctuation marks, and symbols (ie, $, #, @). Stop words (ie, a, the, to) also were removed because they carry little informational content.^35,36 After this process the text consisted of standard English suitable for the R Project for Statistical Computing program (hereafter, R program),³⁷ with which we performed the stemming process, eliminating prefixes and suffixes from different word forms through the identification of word roots to reduce textual volume. For the LDA analysis, the Gibbs sampling method was used with seed words with alpha 0.1.³⁸ The Cao Juan algorithm was applied to achieve a number of topics that was both parsimonious and minimal in terms of overlap.³⁹

After the LDA analysis, all authors had several group meetings to label each LDA-derived topic based on the included words. During the labeling process, we also chose a sample of posts that contained at least 1 of the topic’s terms to examine the consistency between the content of the posts and the label we considered for the topic. If the label reflected the content and all authors reached consensus on the topic’s narrative content, we assigned the label to that topic. We then categorized topics as physically related or emotionally related. For example, if the topic contains understanding, encouragement, or caring, it was categorized as an emotionally related topic.^29,40

RESULTS

Characteristics of participants

The average age of participants was 46 ± 13 years (range = 23-87 years); 35 participants self-identified as female and 2 self-identified as genderqueer (but were assigned female sex at birth). For sexual orientation, 25 participants identified lesbian, 4 as queer, 3 as bisexual, and 2 as gay. Thirty-three participants self-identified as white, 2 as African American, and 2 as Hispanic/Latino. Twenty-four participants were partnered, and, of those, 11 were currently married to their same-sex partner. All participants had been diagnosed with cancer between 1992 and 2016: 29 had breast cancer; 2 had leukemia, thyroid, or ovarian cancers; and 1 each had uterine, vulvar, cervical, kidney, bladder, blood, appendix cancer, osteosarcoma, Hodgkin lymphoma, and non-Hodgkin lymphoma. At the time of the study, 21 were undergoing treatment.

When responding to an open-ended question about why they were seeking support, 8 explicitly mentioned the need to connect with other SMW with cancer, and 6 of them reported having participated in general cancer support groups before (eg, in-person and online support groups).

Characteristics of SMW posts

Each participant wrote 19 ± 43 posts on average (range = 1-255, median = 5) and 52% of participants wrote 5 or fewer posts during their OHC participation (Figure 1). The average number of words per post was 154.2 ± 152.7 (range = 2-2102).

Figure 1.

Histogram of number of posts written by a participant.

Among the 8 OHCs, OHC 1 contained the most active participants, whereas OHC 5 and 7 had the least active participants, despite their large size. In all OHCs, the moderators initiated more threads than the users, with OHC 1 showing the highest percentage of moderator-initiated threads (88%) and OHC 7 the lowest percentage of moderator-initiated threads (54%), indicating that OHC 7 users most actively led communication. See Table 2 for OHCs characteristics.

Table 2.

OHCs by characteristics of posts

	OHC 1	OHC 2	OHC 3	OHC 4	OHC 5	OHC 6	OHC 7	OHC 8
Participants	5	7	10	11	10	6	11	9
Moderator posting	93	71	82	72	38	43	38	50
Participant posting	121	96	191	162	74	79	76	85
Participant-initiated threads	3	10	5	7	9	7	6	2
Moderator-initiated threads	22	19	21	15	14	12	7	13
Total threads	25	29	26	22	23	19	13	15
OHC engagement rate (mean of postings per participant in each group)	24	14	19	15	7	13	7	9

OHC: online health community.

LIWC results

Each post sentence contained 14.7 words on average, and 89% of the words comprising the text of the OHCs were represented in the LIWC dictionary. The LIWC standardized summary scores (range = 0%-100%) were 44% (analytical thinking), 39% (clout), 71% (authenticity), and 58% (emotional tone). LIWC summary scores are considered “high” if the score exceeds 50%. Among the 88 categories, 7.2% of the total number of words in OHCs were identified as emotion words, 8.8% of words were social process related words, and 10.6% were categorized as cognitive processes. Another 3.9% of words in in the OHCs were related to biological processes (for details, see Supplementary Appendix A).

Table 3 shows the statistically significant correlations between OHC engagement and LIWC word categories. Of the 52 categories, sadness, female-reference, drives, and religion were positively and significantly correlated with the OHC engagement rate (P < .05), indicating that higher usage of sadness, female references, drives, and religion-related words were positively associated with writing a higher number of posts during the OHC.

Table 3.

Statistically significant correlations of Linguistic Inquiry and Word Count categories with user online health community engagement

Word category	Pearson correlation coefficient (r)
Sadness (eg, crying, grief, sad)	0.81a
Female-reference (eg, a term referring to a female individual)	0.76a
Drives (eg, an overarching dimension that captures needs and motives of writers)	0.75a
Religion (eg, altar, church)	0.77a

^a P < .05.

Topic modeling

The topic modeling analysis resulted in 10 topics based on the algorithm,³⁹ characterized by 15 words that exhibited the highest association with the topic (Table 4). We labeled the 10 topics: coping, holidays and vacation, cancer diagnosis and treatment, structure of day-to-day life, self-care, loved ones, physical recovery, support systems, body image, and symptom management.

Table 4.

Vocabulary samples from latent Dirichlet allocation topic analysis

Topic	Sample vocabulary	Focus area
Coping	Cancer, know, feel, now, want, think, life, good, things, never, people, see, way, work, many	Emotion
Holidays and vacation	Love, family, hug, new, [individual’s name], great, home, hope, Phoenix, see, good, nice, Christmas, fun, summer	Emotion
Cancer diagnosis and treatment	Cancer, breast, diagnose, surgery, treatment, stage, mastectomy, months, chemo, radiation, live, two, early, name, work	Physical
Structure of day-to-day life	House, horse, week, home, work, last, things, night, farm, snow, working, dad, busy, old, winter	Emotion
Self-care	Eat, diet, good, better, try, take, thanks, lot, meat, exercise, follow, food, [individual’s name], healthy, now	Emotion
Loved ones	Mom, love, lost, friend, daughter, best, friends, [individual’s name], today, life, miss, remember, many, cry, came	Emotion
Physical recovery	Pain, doctor, first, side, surgery, today, better, called, told, went, knee, hours, sleep, need, onc	Physical
Support systems	Group, support, doctor, health, many, people, medical, area, mostly, gay, insurance, social, women, system, though	Emotion
Body image	Scars, breast, body, mastectomy, now, reconstruction, surgery, women, weight, pink, wear, chest, level, used, flat	Physical
Symptom management	Watch, remember, kids, wife, brain, situation, symptom, word, diagnose, movie, despite, end, fatigue, movies, PTSD	Physical

A first name contained in diverse topic’s term lists, was substituted with [individual’s name].

Table 4 lists the 10 topics and the associated words that are most representative of each LDA topic.

Coping was the most common topic and symptom management was the least common topic (Figure 2). Four topics (cancer diagnosis and treatment, body image, symptom management, and physical recovery) were physically focused, whereas 6 topics were focused on emotional or psychological well-being. Not all words were unique to the each LDA topic; some words (eg, cancer, good, life) emerged in multiple topics.

Figure 2.

Frequencies of each topic derived from the latent Dirichlet allocation.

DISCUSSION

The current study is the first, to our knowledge, to examine topics of OHCs designed for SMW with cancer. The OHCs were designed to allow SMW to formulate, reflect on, share, and respond to concerns, feelings, or questions regarding cancer, cancer treatment, and living with cancer, while receiving support from other individuals who shared their sexual identity. We applied LIWC and LDA to understand SMW’s experiences of cancer in greater depth, and to guide researchers and clinicians who wish to address sexual minority cancer survivors’ needs through developing OHCs.

Many of our findings in SMW with cancer appear to be consistent with prior research that focused on OHCs of presumed heterosexual cancer survivors. It has been shown that cancer OHCs function as a channel for survivors not only to exchange health information about their disease, but also disclose their personal stories and seek emotional support from peer survivors.^31,41,42 Hartzler et al⁴³ showed that knowledge and information offered by patients in OHCs were more focused on coping with personal issues in the daily context. Also, Wang et al²⁹ showed that OHC members who provided more emotional support in their postings were more likely to remain in the OHC. The success of peer support as a modality to improve well-being is largely based on this process of sharing experiences, coping strategies, and stories of resilience.^43–45 This study suggests that OHCs are also a valuable modality for SMW with cancer.

We found that participants used more emotional, and psychological words compared with biological words, as has been shown previously in studies on OHCs for heterosexual women with cancers.^46,47 Moreover, higher usage of sadness, female references, drives, and religion-related words was positively and significantly associated with a higher number of posts during the OHCs. Correlations between sadness, drives and OHC engagement of the SMW with cancer were consistent with prior research on OHCs for heterosexual participants with chronic illness.^48,49 In general, people often turn to OHCs when they are struggling with stressful circumstances. This pattern has been shown in heterosexual cancer survivors; sadness is the negative feeling most frequently expressed in OHCs of other chronic diseases, including cancer.⁴⁹ Another finding was a correlation between using religious words and OHC engagement. Generally, religious messages are associated with emotional support,⁵⁰ and a prior study of sexual minority individuals showed that religious coping can moderate minority stress due to sexual minorities’ disadvantaged social status.⁵¹ This study extends previous findings to show that religious coping may be used by SMW with cancer to support their peers in OHCs. Future research should examine the context of religion and spirituality for SMW in coping with cancer.

While LIWC analysis allowed us to examine the textual content of the OHCs at the word level, the LDA approach allowed us to understand the posts at the topic level. Doing so, we found that coping was the most frequently mentioned topic across the OHCs; in contrast, symptom management was the least-mentioned topic. The majority of topics that emerged from these OHCs were related to the psychological and emotional state of survivors, rather than their physical issues. The topics that emerged from the SMW were largely similar to those identified in research of OHCs of heterosexual cancer survivors.^29,30 In fact, many topics directly overlapped between SMW and heterosexual women with cancer, including side effects and support systems.^27,28 Later studies that have a direct comparison of sexual minority specific and generic OHCs for women will be in a better position to identify differences in the meaning of these topics for the 2 groups of women.

Despite this apparent similarity between our findings and the existing research of OHCs derived from heterosexual cancer survivors, we noted unique ways in which SMW processed these topics and related them to each other. In the OHCs exclusively for SMW, highly engaged participants used female-reference words. Using female-reference words means that either the writer is female or the writer refers to other members who are female.⁵² This finding suggests that SMW in OHCs connected to others via their shared female gender identity. Surprisingly, sexual identity came up rarely in the OHCs. Sexual identity related words were included only once across the LDA topics: the term gay in the support systems topic, which is the social identity-related topic. In this case, gay may refer to other SMW, a term in use among older lesbians. The use of gay in the context of support systems suggests that SMW with cancer rely on other sexual minority individuals as support systems, as it has been shown in prior research.^10,28 Additionally, upon close examination of the posts, SMW in the OHCs seldom explicitly touched on sexual minority identity, discrimination, or stigma. SMW shared their sexual identity, but did not explicitly discuss their identity in their posts. We speculate that connecting with other SMW peers and feeling comfortable disclosing personal and private stories and sharing their experiences became possible because the offered OHCs were for SMW only. Future studies that have data to directly compare OHCs for SMW to heterosexual women can identify differences by sexual orientation.

We found that SMW participating in these 8 OHCs tended to write postings when moderators initiated the communication rather than when other members of the OHC initiated. Mostly, the flow of discussion in our OHCs was stimulated and managed by 2 moderators, who frequently asked participants about psychological and emotional aspects of their functioning. Most of the prompt questions were about how participants coped with cancer in their daily lives (eg, “How was your day?”), which likely led to majority of posts being coping-related. Most OHCs for cancer survivors currently run without moderators. However, moderators can create and sustain vibrant online communities by prompting discussion and synthesizing themes.^27,44,45,53 Moderation may be necessary to maintain participant engagement.^41,54,55 In our study, the moderators posted topics when there was a lull in participation, monitored all threads posted on the OHC, and provided responses to participant questions or points of discussion as soon as possible. Building rapport between moderators and participants may be an essential ingredient in promoting participant engagement and creating an empathic community that will have a positive effect on the health of OHC participants. Subsequent studies of this sort must identify strategies for (1) utilizing moderators, (2) encouraging sexual minority participants who are more reticent to engage with the OHC, and (3) utilizing the enthusiasm and presence of very engaged participants to build a sense of community.

Clinical implications

We found that SMW who were more engaged in the OHCs tended to use more female-reference words, which implies connections via sharing female gender identity. Given that coping was the most common topic across all OHCs, SMW may use safe spaces like sexual minority focused OHCs to support one another as women and cope communally with stressors, including the stress of cancer. They can also help each other to cope with the stress of gender-based discrimination, as highlighted by previous literature.^11,56 Many SMW in their demographic questionnaires stated that they wanted a group comprising only SMW, perhaps because of the mutual understanding engendered by a shared SMW identity. Studies of sexual minority cancer survivors have concluded sexual minority-specific support services are needed to provide safe spaces in which survivors can discuss discrimination experiences and heterosexism.⁵⁷ SMW cancer survivors experience discomfort and isolation in support groups made up primarily of heterosexual women,^58–60 drop out of mainstream support groups,^61,62 and may have different support needs in groups.^63,64 Current analysis suggests that clinicians, researchers, and organizations need to consider offering secured OHC platforms to SMW cancer survivors.

Limitations

Because of the relatively small number of participants who took part in this study compared with other studies of OHCs, the generalizability of our results must be viewed with caution. In terms of using data mining to understand the needs of a given participant population, the benefit of any OHC lies in gathering a large number of posts and interactions in a short amount of time. Moreover, our OHC was not a public discussion board—it was available to only sexual minority participants who applied for participation and filled out registration forms. Therefore, participant characteristics may differ from the broader SMW community with cancer. Among our 46 participants, 35 participants self-identified as female, 2 self-identified as genderqueer, and 9 did not specify their gender identity. Individuals of different genders and specific cancer types may have different needs; however, we could not identify those specific needs since the majority of participants were women with breast cancer.

In other studies, LIWC has been used to quantify and categorize words (eg, personal, emotional, sexual) as a linguistic variable to predict behaviors or health outcomes.^64–67 Unfortunately, we were unable to link the posts in these OHCs with any quantitatively derived health outcomes because quantitative data was not linked to the OHCs. Therefore, we used LIWC as a descriptive tool to examine types of words used by the study participants. Our situation is not unusual. Typically, OHCs for cancer survivors (eg, the Cancer Survivor Network, administered by the American Cancer Society) neither collect personal information from participants nor facilitate correlation with health records. Therefore, in subsequent studies, finding ways to link the text of a given OHC to the health status of survivors will be key. Doing so will allow us to leverage the LIWC dictionary to better understand the needs and concerns of OHC users.

CONCLUSION

OHCs have played a critical role in cancer care by providing survivors with emotional and informational support. Although a large number of these OHCs are available, this is the first study to analyze the use of OHCs among SMW with cancer. Despite its limitations, our study sheds new light on these OHCs by identifying unique perspectives of SMW participants from analysis of text. Additionally, this study demonstrates that an OHC can be a good channel through which SMW cancer survivors can process their experiences of living with cancer and support each other. Moving forward, we hope to extend this innovative resource to better understand the experiences and needs of sexual minority cancer survivors, and eventually develop OHC-based interventions to allow survivors to access individually tailored support.

FUNDING

This study was supported by National Cancer Institute grants K07 CA190529 and UG1 CA189961 (CK). The National Cancer Institute had no role in the design or conduct of the study; analysis, or interpretation of data; preparation or review of the article; or the findings and conclusions in this article.

AUTHOR CONTRIBUTIONS

All authors have made substantial contributions to the conception, and conduct of the study; data acquisition, analysis, and interpretation; and/or drafting, editing, and revising the manuscript.

CONFLICT OF INTEREST STATEMENT

None declared.